INTRODUCTION

This guideline summarizes the available data and the quality of the evidence to provide practical
approaches to the diagnosis, management, and prevention of VTE in pregnancy. VTE remains an important
cause of maternal morbidity and mortality in Canada with an overall incidence of DVT and PE of 12.1 per
10 000 and 5.4 per 10 000 pregnancies, respectively.1 VTE occurs at a rate of 5.4 per 10 000 antepartum,
7.2 per 10 000 peripartum, and 4.3 per 10 000 pregnancies postpartum.1 These rates are consistent with
published literature from around the world.2–4 The first and second trimesters of pregnancy convey
similar risks for DVT, with a higher risk in the third trimester and the first 3 weeks postpartum.5,6 PE
occurs more commonly postpartum, decreasing in incidence after the first 6 weeks.3,7 This guideline
sequentially reviews key components in reducing the risk VTE in pregnancy, which include accurate
diagnosis and treatment of DVT and PE, antepartum thromboprophylaxis in appropriate patients,
peripartum management of anticoagulants, and postpartum thromboprophylaxis, and concludes with a
discussion of the use of heparin to prevent adverse pregnancy outcomes. Making decisions about the
management of individual patients can be challenging and complex. Wherever possible, this guideline
attempts to summarize and organize

the existing evidence that supports the recommendations, and it is meant to be complementary to other
international guidelines on this topic.8–15

ACUTE VENOUS THROMBOEMBOLISM IN PREGNANCy

Due to hormonal influences on vascular tone and compressive effects on veins by the enlarging uterus,
DVT in pregnancy generally presents in the lower extremities, with a predisposition for the left leg (70 to
80%).16,17 In contrast to their presentation in non-pregnant patients, DVTs are often isolated to the iliac
and/or femoral vein during pregnancy (61%).18 Consequently diagnostic approaches advocated for use in
non-pregnant patients require modification in pregnancy.8

Diagnosis of VTE in Pregnancy

In non-pregnant patients, diagnostic approaches for VTE use a combination of validated structured clinical
prediction rules with or without the use of D-dimer testing, followed by objective testing with CUS.8
Extrapolating the same approach to pregnancy is difficult because: 1. structured prediction rules have not
been validated in pregnant women, 2. the anatomic presentation of lower extremity DVT in pregnant
women could affect the sensitivity of CUS,18 and 3. current validated D-dimer level cut-off points are of
limited utility.19,20 The potential use of a pregnancy-specific structured prediction rule and pregnancy-
specific D-dimer thresholds has been reported,21–23 but currently neither test should be used alone or
in combination to diagnose or exclude VTE without further validation studies. Our recommended
diagnostic algorithm for DVT in pregnancy is shown in Figure 1. When a pregnant woman presents with a
suspected DVT, she should undergo an ultrasound including direct visualization of the entire proximal
venous system from the iliac to the popliteal vein.24 Doppler studies should be performed at the level of
the iliac vein to ensure that flow is present. Compression manoeuvers should be performed along the
entire venous system from the femoral to the popliteal vein. The sensitivity and negative predictive value
of this method are 90.9% (95% CI 69.4 to 98.4) and 98.9% (95% CI 95.5 to 99.8), respectively.24 Published
evidence is currently insufficient to support the safety of performing a single ultrasound examination in
pregnant women with suspected DVT. Hence, we would recommend repeat testing with CUS and Doppler
imaging as above at least once again over the next 7 days if the initial study is negative. If isolated iliac
vein obstruction (i.e. absence of flow) is suspected on Doppler examination, two options are available: 1.
institute therapeutic anticoagulation followed by repeat CUS in 2 to 3 days, or 2. proceed with MRI. The
option chosen depends on patient preference, availability of expertise, and access to imaging. The
specificity and sensitivity of MRI and the specific technique used to diagnose DVT in pregnancy remains
uncertain.25,26 When PE is suspected clinically, definitive diagnosis requires diagnostic imaging. Several
factors should be considered in the choice of VQ scan or CTA: 1. the maternal and fetal risks associated
with the tests (radiation and contrast agent), 2. the sensitivity of the tests, and 3. their availability. For
both VQ scan and CTA the calculated radiation risk to the fetus is low, with levels below the threshold of
50 mGy

for subsequent childhood malignancy.27–29 The calculated minimum radiation dose to each breast for
an average 60 kg woman is 20 to 35 mGy from CTA and 0.28 mGy from VQ scan.30,31 While little is known
about the long-term effects of radiation exposure to breast tissue during pregnancy, there are data linking
imaging procedures to an increased risk of breast cancer.32 The iodinated contrast agent required for
computed tomographic angiography to diagnosis PE crosses the placenta and can theoretically result in
fetal or neonatal hypothyroidism. However, this risk was not significant in an observational study of over
300 pregnancies.33
In pregnancy the observed sensitivity and negative predictive values of CTA and VQ scan appears to be
high, using clinical outcome as a surrogate measure.34–38 The specificity of a CTA in pregnancy cannot
be ascertained, but studies in non-pregnant patients suggest CTA might be less specific in younger
patients.39 The decision to use CTA or VQ scan is also dictated by local availability and expertise. The CTA
technique used to diagnose PE in nonpregnant patients should be modified as 5% to 36% of scans can be
inadequate in pregnancy due to physiological changes.32,40–42 We currently advocate the use of the VQ
scan as the diagnostic test in pregnancy whenever possible for the reasons listed above. However, if CTA
is used, it is important to counsel patients regarding breast radiation and ensure local awareness for
technique modification.

TREATMENT OF ACUTE VTE

Setting Once an acute VTE is confirmed, therapeutic anticoagulation should be instituted promptly. There
are no studies confirming the safety of outpatient management in pregnancy for women with acute VTE.
Given the additional fetal concerns, pregnant women with an acute PE and/or a large proximal DVT should
be considered for hospitalization or followed closely as outpatients in the initial two weeks following
diagnosis if they remain hemodynamically stable.

Recommendation 5.

Pregnant women diagnosed with acute venous thromboembolism should be hospitalized or followed
closely as outpatients for the first 2 weeks after the initial diagnosis. (III-C)

Choice of anticoagulant

Vitamin K antagonists, such as warfarin, should not be considered for the treatment of VTE in pregnancy
except in exceptional circumstances. They cross the placenta, and first trimester exposure can cause
warfarin embryopathy (midfacial and limb hypoplasia, stippled bone epiphyses).44,45 They are also
associated with pregnancy loss and fetal anticoagulation at the time of delivery.46 UH and LMWH do not
cross the placenta and do not cause teratogenicity or fetal bleeding.47–52 HIT occurs in 3% of non-
pregnant patients receiving UH. It has never been reported in a pregnancy with LMWH,51 and outside of
pregnancy HIT has been reported only in rare cases.53 Due to its lower side-effect profile and ease of
dosing, LMWH is recommended over UH for use in pregnant women. Table 2 outlines the pooled risk
estimates of side effects associated with LMWH use in pregnancy. The specific LMWH preparation used
depends on availability and costs. There is no current evidence to

suggest the superiority of one preparation of LMWH over another. Danaparoid and fondaparinux are
heparanoid molecules that do not cross-react with HIT antibodies. Both are treatment options for
pregnant women with evidence of HIT or allergic reactions to heparins.54,55 These agents should only be
used after consultation with an appropriate specialist. There are currently no data on the safety in
pregnancy of the oral direct thrombin inhibitors (dabigatran) and factor Xa inhibitors (rivaroxaban and
apixaban). Given their very low molecular weights, they are likely to cross the placenta and should be
avoided.

Recommendations 6.
Low molecular weight heparin is the preferred pharmacologic agent over unfractionated heparin for the
treatment of venous thromboembolism in pregnancy. (II-2A) 7. Heparin-induced thrombocytopenia in
pregnant women is extremely rare. Consultation with a hematologist or thrombosis specialist is
recommended to consider the use of heparanoids for treatment of venous thromboembolism if it occurs.
(II-3B) 8. Vitamin K antagonists should only be considered in exceptional circumstances for the treatment
of venous thromboembolism in pregnancy. (II-2A) 9. We recommend against the use of oral Xa inhibitors
and oral direct thrombin inhibitors for the treatment of venous thromboembolism in pregnancy. (III-D)

Anticoagulant dosing and monitoring

Recommended doses for anticoagulation medications are presented in Table 3. The specific LMWH dosing
is as per the manufacturer’s recommendation, based on the woman’s weight at the time of presentation.

There are uncertainties surrounding dosing regimens; the need for monitoring and dose increases with
weight gain associated with therapeutic LMWH use in pregnancy.56 While LMWH is administered as a
single daily dose for non-pregnant patients, twice a day dosing is often used in pregnancy, especially for
the first month when the risk of recurrence is greatest. This practice stems from the altered renal
elimination of LMWH and the impact of weight gain, both of which affect anti-Xa activity in pregnant
women.9,14,50,52,57–59 Hence, for the treatment of acute VTE, especially major proximal VTE and PE,
consideration should be given to initial monitoring of anti-Xa activity, during the first month of treatment
only, to target a level of 0.6 to 1.0 U/mL 4 hours after injection, bearing in mind that target levels will vary
with the LMWH used. However, the cost of the assay, the lack of correlation with clinical events, and the
variability between assays makes the utility of monitoring anti-FXa activity in pregnancy controversial.9 If
UH is selected for initial treatment, it should be administered initially as a bolus followed by a continuous
infusion, using a weight-based nomogram to estimate required doses, and adjusting the infusion to keep
the aPTT at 1.5 to 2.5 times baseline. After initial treatment, a switch to therapeutic subcutaneous LMWH
or UH can be made. If UH is selected, it should be administered subcutaneously twice daily with doses
adjusted to maintain the aPTT at 1.5 to 2.5 times pregnancy baseline at the mid-dosing interval (i.e., 6
hours after the last dose). For women with significant renal impairment (GFR < 30 mL/minute) we
recommend UH over LMWH.

Duration of therapeutic anticoagulation

If an acute VTE is diagnosed early in pregnancy, reducing the anticoagulation intensity after 3 months to
intermediate or prophylactic (low) dose LMWH for the duration of the pregnancy is an option, although
evidence confirming or disputing the safety of this option is unavailable. In the postpartum period, both
LMWH and warfarin can be used.